EP0271670B2 - Method for the detection of corrosion or such - Google Patents

Description

The invention relates to a method according to the preamble of claim 1. In the case of pipelines, there is an automatic non-destructive test for corrosion. Pitting or the like is required. Such errors can be determined by changes in the thickness of the tube wall caused by them.

The eddy current method was proposed for crack testing, in which an alternating electromagnetic field of an excitation coil induces eddy currents in the pipeline wall, which are detected by a sensor coil located at the same location. The method has proven itself in the detection of internal cracks, but errors that grow from the outer wall to the inner wall, such as corrosive wall ablation, can occur. Cracks are difficult or impossible to find due to the generally relatively large thickness of the pipe walls (approx. 20 mm and more) and the low penetration depth of the field. A known device using these methods is not suitable for corrosion detection and has not been used for this purpose.

Furthermore, the use of the leakage flux method has been proposed, in which the pipeline wall is magnetized up to the vicinity of the magnetic saturation by using permanent or electromagnets. The magnetic field emerges locally from the wall due to cracks in the pipeline wall. The detection of the stray field with suitable magnetic field sensors enables the cracks to be detected. This method is also more suitable for the detection of cracks and less for corrosion measurements. In particular, it is not possible to determine the residual wall thickness of pipelines.

Another proposal is based on the electromagnetic coupling of ultrasonic waves into the pipeline wall and measurement of the transit time of the ultrasonic pulses reflected on the rear wall. Although no ultrasound coupling medium is required, the method requires a high power input and has a poor efficiency in converting electromagnetic energy into sound energy, so that the signal-to-noise ratio is too small for the required error detection. Furthermore, the method does not allow wall thickness determination in the event of internal corrosion.

US-A-3 810 384 shows a generic method for the detection of corrosion or the like in pipelines by means of ultrasound, ultrasound signals being transmitted through a pipeline during the passage of a pig and used for carrying out measurements, measurement results are stored and these are carried out after the measurement run be evaluated. This is done with a pig, which is moved through a liquid flowing through a pipeline by attacking the sleeves surrounding the pig body. A rigid holding part for ultrasonic transducers, which are arranged on the circumference of the holding part, is fastened to the housing. Electronics are arranged in the housing body and have a measuring or counting circuit. This is only put into operation when a signal reflected by the inside of the pipeline has been received. This triggers a ramp generator, the voltage which rises with time is a measure of the past time, so that the ramp generator acts as a stopwatch, the running of which is stopped after a reflection from the outer wall or a defect. The voltage then reached as the runtime signal of the ramp generator is proportional to the wall thickness sought. The output signal of the ramp generator only represents the residual catch strength regardless of whether there is external corrosion, internal corrosion or both at the same time. The known method is unable to distinguish between external and internal corrosion.

US-A-4 205 554 shows an apparatus for the detection of defects on an object. A sensor unit is provided which can be moved over the object with a drive unit. The sensors must lie directly on the surface. Only the reflection signals of an error location located inside the object are addressed. An essential feature of the method according to this document is that a weakening characteristic for the reflected beam is specified or simulated and only reflection signals which exceed the curve of the weakening characteristic are detected, buffered and digitally processed.

The invention has for its object to provide an improved method for detecting corrosion or the like in pipelines, which is in particular also able determine the radial location of the corrosion and thus differentiate between external and internal corrosion.

According to the invention, the stated object is achieved by a method of the generic type with the characterizing features of claim 1.

The ultrasonic transit time measurement is carried out according to the pulse reflection method by vertical radiation. The ultrasound pulse triggered by the transmission pulse on the test head passes through an oil feed section formed on the test head, is partly reflected on the inner pipe wall, partly reaches the pipe wall and is reflected on the outer pipe wall. Both reflection pulses run back to the test head. Multiple reflections can be switched off by appropriate windows during the measurement. With this method, two transit times are thus determined for each ultrasound pulse emitted, the first of which indicates the distance to the inner tube wall, while the difference between the two determines the thickness of the inner tube wall. Both impulses can be used to determine whether a weakening of the pipe wall is due to internal or external corrosion.

Due to the timing of the measurements, it is not necessary to continuously record and record all the measured values, but only those in the rhythm of the timing, which initially depends on the desired resolution and, moreover, on the ratio of the measurement section to be traveled and storage capacity. By digitizing the measurement results, the prerequisites are created both to store them in a memory-saving form and, if necessary, to prepare them additionally. By means of intermediate storage and only block-wise final storage, the information can be packed tightly on the final storage medium, namely a mass storage device, as is not the case with continuous storage of received measurement data.

According to the method according to the invention, it is provided that in the case of several measurements carried out over the scope of the pig, these are combined in a measurement data block with further information. It is not necessary that, in the case of sensors distributed over the scope of a measuring pig, the measured data of each individual sensor are provided with distance information. It is sufficient if the measurement results of the sensors scanned in the multiplex method during a scanning period are combined and provided with a total distance information. This also applies in particular if, as provided in the patent application EP-A-255 619, the sensors measuring the circumference are not all on a circumferential line, but are arranged in groups offset from one another in order to record the circumference without gaps; in this case, a correction can be made with regard to the path information for a circumferential line for the sensors not lying on it.

Even pigs that have an eccentric weight distribution so that they largely occupy the same azimuth position can fluctuate around the normal position. For this reason, it is advantageous to further provide that the angular position of the pig in the pipeline is determined and also stored, so that each measured value with regard to the circumference of the pipeline can also be assigned to its correct location. This configuration also saves an eccentric weight distribution if the angular position can be measured over the entire circumference. For this purpose, a 360 degree pendulum potentiometer is preferably provided.

In a further preferred embodiment it can be provided that the route information is measured several times and the most probable is stored. During the distance measurement, to which odometer wheels are generally used, errors can occur due to a wheel slipping or turning freely, for example in the area of the connecting piece. It therefore makes sense to first obtain the route information several times and then to use a sensible one for further processing when the information is different. For this purpose, it is provided that several displacement sensors are provided with a displacement processing unit, in particular the odometer wheels being evenly distributed over the circumference. In the case of two path information, the mean value can in principle be taken, but since the most frequently occurring error is the slip, the measured value which indicates the larger path is preferably used further, since this has the lowest slip error. The selection can also be linked to further criteria, such as the difference between the two route information obtained.

With three route information obtained, several types of evaluation are also possible. It is advantageously provided here that with 3 odometer rolls 2 odometer wheels are selected, the measurement results of which show the smallest difference. To correct the odometer counts, you can basically first Welding points of the pipeline are used. In addition, as far as this is possible, markings, so-called bench markers, can be provided outside the pipe at geodetically precisely known locations. The signals emitted by these are detected and stored by a receiver when the pig moves past. These marker signals correct the odometer values during the evaluation.

When carrying out the measurements, it can preferably be provided that only those measurement results are received and processed which are received within a predetermined time range after a measurement signal has been sent. This eliminates errors that can occur due to multiple reflections when performing the ultrasound measurement. The time intervals can be set according to the lead distance or according to the maximum wall thickness. Both the start and end of the time intervals of the pre and wall thickness runtime can be set separately. This prevents, for example, that the second back wall reflex is used for evaluation. Before the pig run, the orifices must be set so that they cover the entire pipe wall thickness range to be expected.

In addition, it can be provided that a low-frequency alternating electromagnetic field is introduced into the pipe wall and detected with respect to the amplitude at a distance from the introduction point, and the phase shift is measured. In this so-called far-field eddy current method, the low-frequency sinusoidal alternating electromagnetic field generated by the exciter coil with frequencies in the range from 50 to 500 Hz is forwarded via the pipeline wall and by sensors that are at a given distance along the wall, in particular at an axial distance to the exciter coil are detected. This enables sensitive error detection by measuring the phase shift between the sinusoidal signal of the transmitter coil and the sinusoidal signals received by the sensors. The method mentioned can be used in particular for the detection of pitting but also for the detection of cracks. Furthermore, the detection of natural corrosion and weld seams with high sensitivity is possible not only on the inside but also on the outside. Induction coils or Hall generators are preferably used as sensors.

The abovementioned measures with regard to the acquisition of the measurement data mean that, despite the large amount of information that is available, suitable mass storage devices in the gigabyte range can store measurements over a sufficient path length of several 1000 km.

Another embodiment is characterized by the fact that only the difference between corresponding measured values is stored, whereby it can further be provided that if measured value differences occur below a predetermined value, several, in particular two, measured values are each displayed separately in a single byte and in a guide byte the mode and the number of bytes are entered in this mode. The final storage of cached data is generally much less frequent than the intermediate storage. The intermediate storage and only transfer of larger data blocks in the order of MByte to the mass storage means that the data can be stored much more densely on the mass storage than a continuous recording and thus a larger amount of data can be stored.

For evaluation purposes, it is preferably provided that pipe wall defects are displayed according to their distribution over their walls, in particular different depths of the defects being displayed with different colors.

In order to be able to determine the cross section of corrosion or pitting points, a further preferred embodiment provides that cross sections of the tube wall, in particular in areas of flaw points, are shown.

Further preferred refinements provide that measurements and processing of measured values take place only at finite speed, and that no processing of measurements takes place below a predetermined pig speed value.

Fig. 1 a

eine schematische Darstellung einer Meßanordnung bei Ultraschall und

b

Abstand Sensor-Wand bzw. Wandstärke;

Fig. 2 a

eine schematische Darstellung einer Wirbelstrom-Meßanordnung und

b

Meßergebnisse einer Wirbelstrommessung über eine untersuchte Fläche;

Fig. 3 a

eine schematische Verdeutlichung zweier Darstellungsweisen,

b

die sich ergebende Querschnittsdarstellung und

c

eine Messung in flächiger Übersichtsdarstellung;

Fig. 4

ein Blockschattbild der erfindungsgemäßen Einrichtung.

Further advantages and features of the invention result from the claims and from the following description, in which an embodiment of the invention is explained in detail with reference to the drawing. It shows:

Fig. 1 a

is a schematic representation of a measuring arrangement with ultrasound and

b

Distance sensor-wall or wall thickness;

Fig. 2a

is a schematic representation of an eddy current measuring arrangement and

b

Measurement results of an eddy current measurement over an examined area;

Fig. 3 a

a schematic illustration of two modes of representation,

b

the resulting cross-sectional representation and

c

a measurement in a flat overview;

Fig. 4

a block diagram of the device according to the invention.

The measuring and storage device 1 according to the invention is housed, apart from an externally connectable terminal, in the housing of a pig which can be moved through a pipe by pressure difference. The device 1 has an ultrasound measuring system 2 (FIG. 4).

For example, several ultrasound sensors 3 are suitably distributed over the circumference of the measuring pig. Channels of the electronics assigned to the sensors 3 are queried using the multiplex method. An ultrasound signal sent against a wall 4 is first reflected on the surface facing the sensor as a first measurement signal 6, which characterizes the distance of the sensor from the wall, and is reflected on the back of the wall 4 as a further measurement signal, the transit time difference being a measurement signal 7 for the wall thickness t results (Figure 1). In FIG. 1 b, only the measurement signal 7 initially shows a deviation from the normal value and thus the reduction in the wall thickness t at the point designated “outside” in FIG. 1 a, while in this area the distance A of the sensor from the wall indicates Signal 6 shows no change. However, this measurement signal 6 also shows a deviation at the second corrosion point designated “interior” in FIG. 1, as does measurement signal 7. Measurement signal 6 here shows the change in distance A of the sensor wall and thus indicates that internal corrosion is present while the measurement signal 7 in turn indicates the total wall thickness t.

Figure 2a shows schematically the arrangement of an excitation coil 3 'and its associated sensors 3˝ in a pipeline with the wall 4. Figure 2b shows the delay signal obtained, i.e. the Ohner connection between the excitation signal and the sensor signal, the tips 6 ', 7' indicating damaged areas of the tube wall 4. A recording curve gives a longitudinal section through the tube wall (arrow 0 in FIG. 3a), while the family of curves represents the change over the azimuth. FIG. 3 b once again shows a cross-sectional illustration as an example for an ultrasound signal, while FIG. 3 c shows a planar view corresponding to the arrow Or FIG. 3 a. Different wall thicknesses can be represented by different grids or color applications: Area B, for example, shows a deep red deep hole, which is surrounded by a somewhat flatter area C - yellow. In the area C '. C 'are a large number of individual holes of the same depth (yellow), while the area D has flat bumps - blue. Fine gradations e.g. can be chosen by applying paint. The standard wall thickness is represented by the background color. The color codes cannot be reproduced here, but can only be explained in the above manner.

The sub-processes given by measurement, data acquisition, possibly compression of the data, recording and control processes are processed by several separate processing systems assigned to the respective sub-processes, namely data acquisition and compression unit 8, recording computer 9, and master 14 (FIG. 4). The various computer systems communicate with each other via bus or parallel coupling.

The measuring system 2 is followed by a data acquisition unit 8, which can also have a module for compressing the data. The recorded data are transferred to a recording computer 9, which has a buffer. In the data acquisition unit 8, the acquired Measurement data combined with further data, in particular on the location of the measuring pig and its angular position in the tube. For this purpose, a marker unit 11, an angle encoder unit 12 and a displacement encoder unit 13 are provided. In order to relieve the data acquisition unit, these additional data are clocked, acquired by the travel sensor 13 via the recording computer 9 and linked to the other measurement data by the data acquisition computer 8 via a dual-port RAM. The overall control of the device according to the invention is carried out by a master computer 14. In the test phase, the master 14 can call up and display the acquired data from the recording computer 9. After a pig run, the magnetic tape is connected to a personal computer in order to read and evaluate the data.

By means of a counter and timer clocked by the data acquisition computer 8, ultrasound transmitters, for example, are activated in such a way that ultrasound pulses are triggered at intervals of less than 100 »s, for example every 39 or 78» s. The sensors are fired sequentially while maintaining an angular misalignment of approx. 175 °, whereby the least possible mutual influence of the individual sensors is achieved. With 64 sensors distributed over the circumference of the pig and pulse triggering every 78 »s, a total scanning time over the circumference of 5 milliseconds results, so that at an average pig speed of 1 m / sec. the distance between two scanning points in the longitudinal direction is 5 mm. If the pulse sampling time is reduced to 39 »seconds, the pulse-path distance can be shortened to 2.5 mm, so that with a scanning diameter of 6 mm in the longitudinal direction an area-wide scanning can be achieved. At a sampling frequency of 12.8 KHz (corresponding to firing every 78 »s), the two reflected measurement signals mentioned above result in a data rate of 25,600 measured values per second on the front and rear wall surfaces of the pipeline wall. At the pig speed mentioned, this data is obtained for a period of 83 hours in a 300 km long pipeline, for example. The transit time of the reflected signals is digitized using a transit time counter, which is, for example, clocked at 29.6 MHz, which results in resolutions of the ultrasound system of 0.1 mm in the tube wall and 0.021 mm in the space between the sensor and the wall. Since lower resolutions are sufficient, the data can be recorded, for example, with a resolution of 0.2 mm wall thickness and 0.33 in the gap, with a digital display in 8 bits for each measured value having a maximum detectable wall thickness of 51 mm and a distance of 82 mm gives what is completely sufficient.

Overall, the foregoing shows the high data transmission speed and the overall information about the aforementioned exemplary measurement run.

The pulse repetition frequency and thus the number of sensors and the frequency of the total scanning is initially determined by the maximum data rate of the final memory, namely a mass storage device, such as a magnetic tape, at a predetermined desired resolution, which at 1.6 Mbit / sec. can lie.

In order to reduce the data rate and amount of information supplied by the data acquisition unit 8, data compression is preferably carried out. This is done by not storing each measured value separately as such, but instead, after measuring a measured value, in particular the standard values of wall distance and wall thickness of the pipeline, only similar or identical measured values are counted. Similar or identical means that the subsequent values may only differ from the initial value by a certain amount that can be preselected in order to be counted as identical or similar only. All that then needs to be stored is the number of the same or similar measured values until a measured value that exceeds the predetermined limit occurs. The number of the same or similar measured values can be recorded in that the most significant bit (MSB) for representing the measured value is not set in a digital word representing a measured value, for example a byte. This can be set separately if a measured value is the same or similar to a previous measured value. This results in a different interpretation of the word or byte during the evaluation. The number of measured values which have changed by a maximum of a predetermined value in comparison with the last measurement taken are recorded in the least significant bits. If a byte is still used as the word, a maximum of 128 measured values can be counted as the same or similar in this form. If there are fewer measured values, the most significant bit is not set in a subsequent byte, so that the following byte is again interpreted as a measured value.

In order to be able to pig longer lines with a large number of sensors, the compression factor can be improved further. The "Difference less than 16" mode is also used. If the difference between the measured values is less than 16 (decimal), several, in particular two, measured values can be differentiated in one byte. The mode and the number of bytes in this mode are entered in a guide byte gene.

In addition to the depth of the corrosion points, the location and their extent are also of interest in the evaluation. On the other hand, the pig is pushed through the pipeline by the medium or the pressure difference before and after it, so that its speed can change. A distance measurement is therefore carried out and measurement data and distance are assigned to one another. For this purpose, a displacement sensor unit 13 or an odometer system is provided, which receives distance impulses from several odometer wheels. Since errors can occur due to slip effects or free-running odometer wheels within T-pieces, it has proven to be advantageous to form the mean value of the adjacent distances for several odometer wheels. This can minimize errors. To mark the path, electronic markings can be provided along the pipeline or placed while the measuring pig is running. The signals emitted by these markings are detected by the marking unit 11 when the pig is passed and fed to the recording computer 9.

For a clear definition of the azimuthal position of the pig, which is necessary to localize the corrosion over the circumference of the pipe, a 360 ° pendulum potentiometer is used to determine an analog-digital converter.

The master system 14 monitors the other electronic components, being programmed by the travel sensor 13 depending on the path. The voltage supply of the master is preferably independent of the supply of the other electronics, so that it can carry out the monitoring of the voltage supply of the other components. A non-volatile memory is provided for storing occurring errors.

For synchronization, i.e. To ensure the retrievability of the individual measurement data, in particular in connection with the distance covered up to their occurrence, the data of the marker 11, the pendulum 12, the travel sensor 13 as well as measured temperature data and the data of the measurement system 2 are measured within a measurement period, i.e. the measured data recorded during a total scanning time of all measuring points distributed over the circumference of the pig are combined with measured values of the front and on the rear wall surface to form data blocks, wherein the individual blocks can be identified. Three bytes are provided for the route information, so that with a sufficient resolution of 10 cm, for example, 1,677 km can be recorded. Regardless of whether the data has been compressed or not, it is transmitted from the data acquisition unit 8 to the recording computer 9, the coupling being carried out via a dual-port RAM as the interface between data acquisition and recording. After being provided with an identifier, the data is first temporarily stored in a buffer of, for example, 1 Mbyte and from there by direct memory access (DMA) in the form of transfer blocks with a size of, for example, 512 K byte to the large mass storage device, such as a magnetic tape with 40 gigabits at one Transmission rate of, for example, 400 kbits, transmitted synchronously in series. The latter is only limited by the maximum recording rate of the mass storage device 17, which is approximately 1.6 MBits per second in the tape device used.

Apart from the preprocessing of the measurement data before the final recording, the data are also not transferred continuously to the magnetic tape for reasons of utilization of the capacity, but rather in blocks. The start-stop operation creates undefined data areas on the tape in the start-up and run-down phases. The transmission block transferred in each case is 512 KB.

After the run has been carried out, the saved data can be evaluated. A first embodiment of the evaluation includes displaying only those measured values whose amounts are within a predeterminable size range. For the ultrasonic measurement, this means that values are displayed that are below a certain minimum wall thickness. In the case of the eddy current method, only those phase shifts that are greater than a minimum phase shift are output.

In a further detailed evaluation, the surface distribution of the measured corrosion spots on the inner surface of the pipe is shown, whereby different corrosion depths can be provided with different color values, on the other hand, representations of corrosion and crack areas of interest are shown graphically in longitudinal section.

Claims (12)

1. Method for the detection of corrosion or the like in pipelines by means of ultrasonics, ultrasonic signals being emitted during the travel of a scraper through a pipeline, the transit time difference between the signals reflected on the inner and outer walls being measured, measuring results being provided with a distance information, stored and the measuring results are evaluated after performing the test run, characterized in that cyclically both the time up to receiving the signals reflected by the inside of the pipeline and the time up to the receiving of the signals reflected by the outside are measured and the difference between the transit times is determined and that these three measuring results are digitized, intermediately stored and blockwise finally stored as measuring data which are compressed in such a way that of a measuring value only following measuring values diverging in a predetermined range are counted and are acquired and stored as a function of their number.
2. Method according to claim 1, characterized in that in the case of several measurements performed over the circumference of the scraper, they are combined into a measuring data block with further informations and the measuring data block is intermediately stored.
3. Method according to one of the preceding claims, characterized in that the distance information is multiply measured and the most probable informations stored.
4. Method according to claim 3, characterized in that in the case of three odometer rolls, two odometer wheels are selected which have the measuring results with the smallest difference.
5. Method according to one of the preceding claims, characterized in that only those measuring results are recorded and further processed, which are received within a predetermined time range following the emission of a measuring signal.
6. Method according to one of the claims 1 to 5, characterized in that a low frequency, electromagnetic alternating field is introduced into the pipe wall, is detected at a distance from the introduction point and the phase displacement is measured.
7. Method according to one of the claims 1 to 6, characterized in that only measuring values corresponding to the difference are stored.
8. Method according to one of the claims 6 to 7, characterized in that use is made of one and preferably the most significant bit of a digital value, particularly byte, for identifying the information type contained in the value (byte).
9. Method according to claim 7, characterized in that when measuring value differences occur below a predetermined value several and in particular two measuring values are separately represented in a single byte and in a guide byte are entered the mode and the number of bytes in the said mode.
10. Method according to one of the preceding claims, characterized in that pipe wall defects are represented as a function of their distribution over the wall.
11. Method according to one of the preceding claims, characterized in that cross-sections of the pipe wall, particularly in the vicinity of the defects are represented.
12. Method according to one of the preceding claims, characterized in that no processing of measurements takes place under a predetermined scraper speed value.

Images